In the field of thermal engineering, one of the biggest concerns is the cooling of heat producing systems. For this purpose, today’s world is encouraging to use such cooling systems which are free from any active components (passive systems) for its high reliability and compact size. For this reason, to establish cooling by transferring heat from one place (source) to another (sink) passive system like natural circulation loop (NCL) is highly used. Fluid flow dynamics of the NCL is changing with the increase in heater power which is used as the source for the simulation. We found steady flow dynamics for the comparatively low power of heat, and with the rise in the power first, we saw the oscillatory flow dynamics and then found flow reversal characteristics. This paper presents a novel strategy for the early prediction of flow reversal phenomenon in NCL using symbolic analysis of time series data. This time series data is found from the numerical simulation, and for the proper study, we are considering data after the initial transient part is overcome. Total time series data is transformed into a symbol string by partitioning into a finite number of specified symbolised groups. The state probability vector is calculated based on the number of occurrences of each symbol group. Present work is a single-phase study, and according to our geometry, we can provide a maximum 800 W heater power to stay in the single-phase. Therefore, for the early prediction of flow reversal in NCL, state probability vector evaluated at 800 W heater power which is the most undesirable state (chaotic data), and this is considered as the reference vector. The difference of the reference state vector from the current state vector is used as a parameter for early detection of flow reversal. It can be observed from the results that this difference changes significantly when the system is sufficiently away from the flow reversal.

在热工程领域，最大的担忧之一是发热系统的冷却。为此，当今世界鼓励使用这样的冷却系统，因为它具有高可靠性和紧凑的尺寸，因此没有任何有源组件（无源系统）。因此，非常需要通过将热量从一个地方（源）传递到另一个（接收器）被动系统（例如自然循环回路（NCL））来建立冷却。NCL的流体流动动力学随加热器功率的增加而变化，加热器功率用作模拟源。我们发现热量相对较低的功率具有稳定的流动动力学，随着功率的增加，我们先看到了振荡的流体动力学，然后发现了逆流特性。本文提出了一种新的策略，用于通过时间序列数据的符号分析来早期预测NCL中的逆流现象。该时间序列数据是从数值模拟中找到的，为了进行适当的研究，我们正在考虑克服初始瞬态部分之后的数据。通过将整个时间序列数据划分为有限数量的指定符号化组，可以将其转换为符号字符串。基于每个符号组的出现次数来计算状态概率向量。目前的工作是单相研究，根据我们的几何形状，我们可以提供最大800 W的加热器功率以保持在单相中。因此，为了早期预测NCL中的逆流，在800 W加热器功率下评估状态概率矢量是最不希望的状态（混沌数据），并将其视为参考向量。参考状态向量与当前状态向量的差用作用于早期检测逆流的参数。从结果可以看出，当系统足够远离逆流时，此差异将发生显着变化。